Increased advanced glycation end products in atherosclerotic lesions of patients with end-stage renal disease.

Abstract

Although advanced glycation end products (AGEs) are increased in the serum and tissues of patients with end-stage renal disease, little is known about the role of AGEs in atherogenesis. We therefore carried out an immunohistochemical study on the accumulation of AGEs and apolipoprotein B in the human aortas of diabetic and nondiabetic subjects with end-stage renal disease. The atherosclerotic lesions included diffuse intimal thickening, fatty streaks and atherosclerotic plaque. We used antibodies against two different epitopes of AGE structures, i.e. an Nepsilon-(carboxymethyl)lysine-protein adduct (CML) and a structure(s) other than CML (nonCML). The area that was positive for an antigen as a percentage of the total area (%Ar) was determined morphometrically, using an NIH-image program. In diffuse intimal thickening, atherosclerotic plaque and tunica media, the %Ar of CML and nonCML was significantly greater in diabetic or nondiabetic subjects with end-stage renal disease than in control subjects without end-stage renal disease. In fatty streaks, the %Ar of nonCML was significantly greater in nondiabetic subjects with end-stage renal disease than in control subjects, while no difference in the %Ar of CML was found between the subjects with or without end-stage renal disease. Nondiabetic subjects with end-stage renal disease showed a significantly increased %Ar of apolipoprotein B in fatty streaks and atherosclerotic plaque than the control subjects. The %Ar of CML and nonCML significantly correlated with the duration of hemodialysis in diffuse intimal thickening and atherosclerotic plaque of subjects with end-stage renal disease, but not in fatty streaks. On the other hand, the %Ar was not related to the duration of diabetes in any of the lesions in the diabetic subjects with end-stage renal disease. In diffuse intimal thickening and atherosclerotic plaque, subjects with end-stage renal disease showed a significant correlation between the %Ar of apolipoprotein B and AGEs (CML and nonCML), as well as their immunohistochemical colocalization. These results suggest that impaired AGE clearance may cause the increased accumulation of AGEs in the aortic wall of subjects with end-stage renal disease, thus resulting in the rapid progression of atherosclerosis. The accumulation of AGEs may be related to an enhanced LDL deposition in atherosclerotic lesions of subjects with end-stage renal disease.

Full text

Page 1

Atherosclerosis 142 (1999) 67–77
Increased advanced glycation end products in atherosclerotic lesions of
patients with end-stage renal disease
Noriyuki Sakataa,*, Yoshinobu Imanagaa, Jing Menga, Yutaka Tachikawaa,
Shigeo Takebayashia, Ryoji Nagaib, Seikoh Horiuchib
aDepartment of Pathology, School of Medicine, Fukuoka Uni?ersity, 45-1, 7-chome Nanakuma, Jonan-ku, Fukuoka 814-0133, Japan
bDepartment of Biochemistry, Kumamoto Uni?ersity School of Medicine, Kumamoto 860-0811, Japan
Received 23 February 1998; received in revised form 8 June 1998; accepted 30 June 1998
Abstract
Although advanced glycation end products (AGEs) are increased in the serum and tissues of patients with end-stage renal
disease, little is known about the role of AGEs in atherogenesis. We therefore carried out an immunohistochemical study on the
accumulation of AGEs and apolipoprotein B in the human aortas of diabetic and nondiabetic subjects with end-stage renal
disease. The atherosclerotic lesions included diffuse intimal thickening, fatty streaks and atherosclerotic plaque. We used
antibodies against two different epitopes of AGE structures, i.e. an N?-(carboxymethyl)lysine-protein adduct (CML) and a
structure(s) other than CML (nonCML). The area that was positive for an antigen as a percentage of the total area (%Ar) was
determined morphometrically, using an NIH-image program. In diffuse intimal thickening, atherosclerotic plaque and tunica
media, the %Ar of CML and nonCML was significantly greater in diabetic or nondiabetic subjects with end-stage renal disease
than in control subjects without end-stage renal disease. In fatty streaks, the %Ar of nonCML was significantly greater in
nondiabetic subjects with end-stage renal disease than in control subjects, while no difference in the %Ar of CML was found
between the subjects with or without end-stage renal disease. Nondiabetic subjects with end-stage renal disease showed a
significantly increased %Ar of apolipoprotein B in fatty streaks and atherosclerotic plaque than the control subjects. The %Ar of
CML and nonCML significantly correlated with the duration of hemodialysis in diffuse intimal thickening and atherosclerotic
plaque of subjects with end-stage renal disease, but not in fatty streaks. On the other hand, the %Ar was not related to the
duration of diabetes in any of the lesions in the diabetic subjects with end-stage renal disease. In diffuse intimal thickening and
atherosclerotic plaque, subjects with end-stage renal disease showed a significant correlation between the %Ar of apolipoprotein
B and AGEs (CML and nonCML), as well as their immunohistochemical colocalization. These results suggest that impaired AGE
clearance may cause the increased accumulation of AGEs in the aortic wall of subjects with end-stage renal disease, thus resulting
in the rapid progression of atherosclerosis. The accumulation of AGEs may be related to an enhanced LDL deposition in
atherosclerotic lesions of subjects with end-stage renal disease. © 1999 Elsevier Science Ireland Ltd. All rights reserved.
Keywords: Atherogenesis; Advanced glycation end products; Apolipoprotein B; End-stage renal disease; Morphometry; Immuno-
histochemistry
1. Introduction
Although atherosclerosis is one of the major compli-
cations of patients with end-stage renal disease, the
molecular pathogenesis of this complication remains
unclear [1,2]. Advanced glycation is a major pathway
for the posttranslational modification of tissue proteins,
and its products are called advanced glycation end
products (AGEs). Modification of AGEs has been
shown to cause functional and structural changes in the
properties of such native materials as enzyme and ma-
* Corresponding author. Tel: +92-801-1011, ext 3281; fax: +92-
863-8383.
0021-9150/99/$ - see front matter © 1999 Elsevier Science Ireland Ltd. All rights reserved.
PII: S0021-9150(98)00192-0

Page 2

N. Sakata et al. / Atherosclerosis 142 (1999) 67–77 68
trix proteins [3,4]. AGEs, including pentosidine and
N?-(carboxymethyl)lysine (CML), have been shown to
be increased in the serum and matrix proteins of pa-
tients undergoing hemodialysis due to end-stage renal
disease [5,6]. Previous studies [7,8] have shown that the
accumulation of AGEs in the artery contributes to the
development of human atherosclerosis. However, little
is known about whether AGE accumulation is in-
creased in the atherosclerotic lesions of patients with
end-stage renal disease.
AGEs that have been isolated from in vivo AGE
proteins include 2-furoyl-4(5)-(2-furanyl)-1H-imidazole,
pentosidine, pyrraline and CML [9–12]. Some contro-
versy remains in the literature concerning the localiza-
tionofAGEs inatherosclerotic
investigators have reported the extracellular formation
of AGEs in atherosclerotic lesions [13], whereas other
data suggest the intracellular accumulation of AGEs in
macrophage/foam cells and smooth muscle cells [8]. We
recently showed that the heterogeneous localization of
AGEs in atherosclerotic lesions depends on their differ-
ent epitopes [14]. In the present study, we therefore
carried out an immunohistochemical analysis, using
antisera against different epitopes of AGEs, consisting
of one reactive to CML and another unreactive to
CML (nonCML).
Low density lipoprotein (LDL) has been implicated
in the pathogenesis of atherosclerosis [15]. Lipoprotein
metabolism has been shown to be disturbed in hemodi-
alysis patients, and to contribute to the atherogenesis of
chronic renal failure [16]. Brownlee et al. showed that
low density lipoprotein was specifically bound by the
advanced nonenzymatic glycosylation products of col-
lagen [17]. Therefore, AGEs may enhance the entrap-
ment of LDL in the arterial wall, and contribute to the
accelerated development of atherosclerosis in patients
with end-stage renal disease.
In the present study, we carried out an immunohisto-
chemical analysis with antibodies against CML, non-
CML andapolipoprotein
accumulation of AGEs and LDL and the relationship
between AGE and LDL deposition in the atheroscle-
rotic lesions of patients with end-stage renal disease.
lesions.Some
B to examinethe
2. Materials and methods
2.1. Subjects and samples
The subjects included six patients with end-stage
renal disease due to diabetic nephropathy (Group 1 in
Table 1), eight patients with end-stage renal disease due
to other renal diseases (Group 2) and eight age-
matched patients without end-stage renal disease as
controls (Group 3). The autopsy was carried out within
2–6 h postmortem. The aorta was slit open longitudi-
nally and the intima was washed with 0.01 M phos-
phate-buffered saline and was then examined by the
naked eye. The lesions were classified as diffuse intimal
thickening, fatty streaks or atherosclerotic plaque. We
did not study complicated lesions that were extensively
ulcerated, calcified or covered by thrombus. The sam-
ples were embedded in OCT compound and rapidly
frozen in liquid nitrogen. Serial frozen sections were cut
on a cryostat and stored at −80°C until use. The
sections were stained with hematoxylin and eosin.
2.2. Immunohistochemistry
Serial tissue sections on microscope slides were fixed
in cold ethanol for 10 min and then immunostained
with an avidin/biotin/alkaline phosphatase system.
Briefly, the sections were incubated for 5 min with 6%
BSA to block nonspecific binding and then reacted with
the primary antibody at room temperature for 60 min.
For the immunostaining of AGEs, we used two pri-
mary antibodies against different epitopes of AGEs,
consistingofonethat
boxymethyl)lysine (CML) and another that was unreac-
tive to CML (nonCML). The specificity of these
antibodies has been tested by ELISA in a previous
study [14]. The anti-? smooth muscle actin (DAKO,
Carpinteria, CA), anti-human macrophage (HAM 56,
DAKO) and anti apolipoprotein B (MONOSAN,
Netherlands) antibodies were purchased from their re-
spective suppliers. After additional washes (in PBS,
three times), biotinylated second antibody (1:200,
Chemicon, CA) was added for 30 min at room temper-
ature. The slides were then rinsed with PBS and incu-
bated for 30 min with alkaline phosphatase-labeled
streptavidin (1:100, DAKO) at room temperature. The
alkaline phosphatase coloring reaction was mediated by
incubation with naphthol-AS-BI-phosphoric acid as a
substrate and hexazotized new fuchsin as a coupler in
0.2 M Tris–HCl buffer (pH 8.2). The reaction mixture
contained levamisole (24 mg/ml) to block the endoge-
nous alkaline phosphatase activity. After washing with
PBS, the sections were counterstained with Meyer he-
matoxylin. Control tests for the specificity of immunos-
taining included the substitution of non-immune sera or
PBS for the primary antibodies. To further verify the
specificity of the staining of monoclonal anti-CML and
polyclonal anti-nonCML antibodies, controls were per-
formed by the preadsorption of anti-AGEs and anti-
CMLfromtheseprimary
AGE-BSAandCML-BSA,
atherosclerotic lesions, e.g. diffuse intimal thickening,
fatty streaks and atherosclerotic plaque, were examined
in each case under a light microscope. Complicated
lesions with calcification and ulceration were excluded
from this study. We examined 147 lesions from 22
aortas, consisting of 51 diffuse intimal thickenings, 48
fatty streaks and 48 atherosclerotic plaques.
was reactive to
N?-(car-
antisera
respectively.
with purified
Three

Page 3

N. Sakata et al. / Atherosclerosis 142 (1999) 67–7769
Table 1
Characteristics of diabetic subjects with end-stage renal disease (Group 1), nondiabetic subjects with end-stage renal disease (Group 2) and control subjects (Group 3)a
TC (mg/dl)
TG (mg/dl)
DM
Duration of DM (years)
No.
Duration of HD (years)
Age (years)
Sex
BUN (mg/dl)
Cr (mg/dl)
FBS (mg/dl)
Group 1 (ESRD+, DM+)
235
78
IDDM
M
28
100
4.3
55
1
7.8
214
7.2
321
150
170
NIDDM
15
–
57
2
M
107
267
206
NIDDM
14
–
203
58
3
7.8
74
M
227
64
114
140
NIDDM
15
1
M
61
6.3
4
9.0
102
96
48
NIDDM
43
14
5
M
73
61
83
74
NIDDM
10
2.5
161
M
6
84
5.6
77
FBS (mg/dl)
Age (years)
TC (mg/dl)
TG (mg/dl)
Renal disease
Duration of HD (years)
Sex
BUN (mg/dl)
Cr (mg/dl)
No.
Group 2 (ESRD+, DM−)
189
1
118
47
Nephrosclerosis
4.83
M
93
16.1
90
145
101
Chronic pyelonephritis
11
86
63
2
6.8
46
M
78
66
265
730
Chronic glomerulonephritis
16
M
37
6.1
3
69
71
73
87
IgA nephropathy
8
M
51
6.4
4
151
75
Chronic glomerulonephritis
12
ND
3.1
5
71
F
19
130
150
Amyloidosis, AL type
14
6
79
F
53
5.7
100
132
110
Hemolytic uremic syndrome
3.4
53
7
4.5
79
F
79
157
8
73
83
Chronic glomerulonephritis
4
M
72
8.1
109
TC (mg/dl)
TG (mg/dl)
Pathological diagnosis
BUN (mg/dl)
Sex
Cr (mg/dl)
No
Age (years)
FBS (mg/dl)
Group 3 (ESRD−, DM−)
0.6
98
180
258
Gastric carcinoma
1
M
47
22
2
232
52
73
Acute pancreatitis, purulent peritonitis
F
14
0.7
126
209
83
Peritoneal serous adenocarcinoma
79
11
F
0.5
3
56
102
63
210
54
Acute myocardial infarction
F
17
0.6
4
0.9
84
245
158
Gastric carcinoma
5
68
M
16
6
132
73
83
Pulmonary carcinoma
M
11
0.8
90
111
114
Liver cirrhosis, esophageal varix
186
7
0.6
75
F
14
8
174
80
74
Pancreatic carcinoma
M
18
1.1
80
aESRD, end-stage renal disease; DM, diabetes mellitus; BUN, blood urea nitrogen; Cr, serum creatinine; TC, plasma total cholesterol; TG, plasma triglycerides; HD, hemodialysis.

Page 4

N. Sakata et al. / Atherosclerosis 142 (1999) 67–77 70
2.3. Morphometry
To evaluate the degree of staining for AGEs and
apolipoprotein B in each lesion, morphometry was
performed using an NIH-image program (Version
1.58). A Nikon microscope equipped with a Hama-
matsu CCD video camera was used to transfer the light
microscopic images to a Power Macintosh personal
computer (8100/110). As shown in Fig. 1, the percent-
age of area that was positive for antigen (%Ar) was
calculated by the following equation: %Ar=100×S/T
(%), where S indicates the area (pixels) that is positive
for antigen and T indicates the total area (pixels)
examined.
To test the reproducibility of the area positive for
CML, nonCML and apolipoprotein B, we calculated
the coefficients of variation for ten samples which were
chosen at random, which were 0.12?0.05, 0.11?0.03
and 0.09?0.01 in CML, nonCML and apolipoprotein
B, respectively.
2.4. Statistical analysis
The dataareexpressed as means?S.D.The
Kruskal–Wallis test was used to compare variables
between the three groups. Mann–Whitney U test was
used to test differences between two groups. Spear-
man’s correlation coefficient was used to identify corre-
lations between: (a) age, duration of hemodialysis or
diabetes mellitus, and the degree of staining of CML,
nonCML and apolipoprotein B; and (b) the degree of
staining of CML, nonCML and apolipoprotein B in
each lesion. The statistical analysis was performed using
commercially available statistical software. A P-value
of less than 0.05 was taken as the level of significance.
3. Results
3.1. Subjects
The clinical characteristics and pathological diagnosis
are shown in Table 1. The mean ages were 65.1?11.3,
69.8?11.5 and 64.2?11.7 years in Groups 1, 2 and 3,
respectively. The mean duration of diabetes mellitus
was 20.8?12.4 years in Group 1. The duration of
hemodialysis was significantly longer in Group 2
(9.15?4.80 years) than in Group 1 (3.63?5.33 years).
There were no differences in the levels of plasma total
cholesterol and triglyceride among the three groups.
3.2. Immunohistochemical localization of AGEs and
apolipoprotein B in aorta
Fig. 2 shows the immunohistochemical localization
of nonCML and apolipoprotein B in diffuse intimal
thickening. In control subjects, scattered staining of
CML and nonCML was observed in the collagen fibers
of intima of older subjects (Fig. 2A), but these were
scarce in those of middle-aged men. Apolipoprotein B
was accumulated not only in AGE-positive collagen,
but also in other extracellular matrix of intima (Fig.
2B). On the other hand, the deposition of CML and
nonCML was increased in the extracellular matrix of
intima in subjects with end-stage renal disease, com-
pared to control subjects (Fig. 2C), where apolipo-
protein B was colocalized (Fig. 2D). In fatty streaks, all
of the groups exhibited numerous macrophages/foam
cells that showed marked CML accumulation (Fig. 3B,
D). Whereas nonCML was deposited in the extracellu-
larmatrix ofintima, it
macrophages/foam cells (Fig. 3A, C). The fatty streak
lesions of subjects with end-stage renal disease showed
a slight increase in the deposition of CML and non-
CML in the extracellular matrix, but no differences
were observed in the degree of staining of CML in the
macrophages/foam cells among the three groups. Fig. 4
shows immunohistochemical staining for CML, non-
CML and apolipoprotein B in atherosclerotic plaques.
The intracellular accumulation of CML was observed
in the macrophages/foam cells in atherosclerotic plaque
of all of the groups (Fig. 4B, E). In some of the control
subjects, the extracellular deposition of CML appeared
in the central atheroma of atherosclerotic plaque. Im-
munoreactivity for nonCML was homogeneously found
in the acellular fibrous area and central atheroma of all
of the groups, but not in macrophages/foam cells (Fig.
4A, D). As shown in Fig. 4D and E, the atherosclerotic
plaque of subjects with end-stage renal disease showed
a marked increase in immunoreactivity for CML and
nonCML in the extracellular matrix compared to that
of control subjects. Apolipoprotein B co-existed with
nonCML or CML in the extracellular space of
atherosclerotic plaque in each group (Fig. 4C, F). Im-
munoreactivity for CML in media is shown in Fig. 5.
was notfound within
Fig. 1. Representative morphometric analysis of the degree of stain-
ing of AGEs in atherosclerotic plaques. (A) Light microscopic image
of immunohistochemical staining of nonCML. I, tunica intima. (B)
Morphometric analysis using an NIH-image program. S, area posi-
tive for nonCML; T, total area of tunica intima.
Fig. 2. Comparative immunohistochemistry of diffuse intimal thick-
ening from subjects with and without end-stage renal disease. (A, B)
Diffuse intimal thickening from a 75-year-old subject in Group 3
(control subjects). (C, D) Diffuse intimal thickening from a 47-year-
old subject in Group 2 (nondiabetic subjects with end-stage renal
disease). (A, C) nonCML; (B, D) apolipoprotein B. Original magnifi-
cation ×400.
Fig. 3. Comparative immunohistochemistry of fatty streaks from
subjects with and without end-stage renal disease. (A, B) Fatty
streaks from an 80-year-old subject in Group 3 (control subjects). (C,
D) Fatty streaks from a 57-year-old subject in Group 1 (diabetic
subjects with end-stage renal disease). (A, C) nonCML; (B, D) CML.
Original magnification ×150.

Page 5

N. Sakata et al. / Atherosclerosis 142 (1999) 67–7771